***********************************************************************************************
					Replication files for: 

			"More Important, but For What? The Insignificant Role
			of Subjective Importance in Vote Decisions"

			Thomas J. Leeper, Associate Professor		
			Joshua Robison, Post-Doc, Aarhus University (jrobison@ps.au.dk)
			

**************************************************************************************************

This repository includes STATA .do files needed to recreate the ANES Time Series and Panel analyses as well as the data and .do files needed to recreate the experimental analyses reported in the manuscript. 

The data files for the ANES can be accessed via the ANES' Data Center: https://electionstudies.org/data-center/

The .do files are of two types: Data Cleaning and Data Analysis files. The former 'clean' the data in preparation for further analysis. The Data Analysis files, meanwhile, contain the code relating to...the analyses. One must run the cleaning files first to produce the analyses. This can be done from within the Data Analysis do files as they will begin with code that runs the cleaning files. One just has to update the file directory in which these files are contained. 

Below is a breakdown describing which files are necessary for each analysis reported in the paper and supplementary materials: 

***ANES Time Series Analyses (Figures 1 and 2; Online Appenndix B; Figure OC2 in Online Appendix C)

Data files for the 1980, 1984, 1996, 2000, 2004, 2008 ANES Time Series and 2008 ANES Panel data (see above). 

Data Cleaning - 1980 ANES TS
Data Cleaning - 1984 ANES TS
Data Cleaning - 1996 ANES TS
Data Cleaning - 2000 ANES TS
Data Cleaning - 2004 ANES TS
Data Cleaning - 2008 ANES TS
Data Cleaning - 2008 ANES Panel
Data Analysis - ANES Time Series Analyses


***Experiment Analyses (Figure 4, Online Appendix A, and part of Online Appendix C)

Conjoint Experiment Data.dta

Data Analysis - Conjoint Experiment (Figure 4 and Online Appendix A)
	note: the above do file focuses on the analyses wherein we reshape the data file so that each row is a respondent/vignette pairing. This corresponds to our in-text 			analyses. We also provide code for the panel attrition analyses reported in Online Appendix A in this file.  

Data Analysis - Conjoint Experiment (Fully Stacked Data)
	note: the above .do file focuses on analyses wherein fully stack the data so that each row is a respondent/CANDIDATE pairing. Such analyses are reported in Figures OA8 and OA9 and Table OA6 and Figure OA12. 


****Online Appendix D
Data files for the 1980, 1984, 1996, 2000, 2004, 2008 ANES Time Series and 2008 ANES Panel data (see above). 

Data Cleaning - 1980 ANES TS
Data Cleaning - 1984 ANES TS
Data Cleaning - 1996 ANES TS
Data Cleaning - 2000 ANES TS
Data Cleaning - 2004 ANES TS
Data Cleaning - 2008 ANES TS
Data Cleaning - 2008 ANES Panel
Data Analysis - ANES Time Series (Online Appendix D)


****Online Appendix E
Data files for the 1980, 1984, 1996, 2000, 2004, 2008 ANES Time Series and 2008 ANES Panel data (see above). 

Data Cleaning - 1980 ANES TS
Data Cleaning - 1984 ANES TS
Data Cleaning - 1996 ANES TS
Data Cleaning - 2000 ANES TS
Data Cleaning - 2004 ANES TS
Data Cleaning - 2008 ANES TS
Data Cleaning - 2008 ANES Panel
Data Analysis - ANES Time Series (Online Appendix E)